Le 4 mai 2021, la plateforme Yahoo Questions/Réponses fermera. Elle est désormais accessible en mode lecture seule. Aucune modification ne sera apportée aux autres sites ou services Yahoo, ni à votre compte Yahoo. Vous trouverez plus d’informations sur l'arrêt de Yahoo Questions/Réponses et sur le téléchargement de vos données sur cette page d'aide.
Question about the t-statistic?
How can the t-statistic be interpreted? (I mean is there a meaning to the absolute value of the T?, I already know it's used to check if the means of two samples are identical)
@ niwriffej:
I'm glad to know you know the exact terms to describe the t-test, I would be even gladder if you answered my question
Maybe I wasn't clear enough...I already know what a t-test is and when to use it, my question is about the t-statistic: does it have a meaning? Can it be interpreted in any way? I found definitions that said it was "a measure of how extreme a statistical estimate is", which doesn't tell me much.
2 réponses
- cvandy2Lv 6il y a 1 décennieRéponse favorite
The t-statistic is used exactly like the Z-statistic (standard normal variate) except it's used for analyzing experiments with sample sizes less than 30. The t-statistic distribution is symmetrical and bell-shaped like the normal distribution except it's a little wider. For samples sizes < 30 you don't have as good an estimate of the population standard deviation as you do with a larger sample - that's why you would use the t-statistic.
It's used not only for comparing 2 population means, but also for comparing a population mean against a standard value and also for confidence intervals.
- niwriffejLv 6il y a 1 décennie
It's not the identical part that matters, it's whether or not the two samples have statistically signifigant differences. That's primarily what the t-test is for.